Düngupeyem (Morphological analyser and generator for Mapudüngun)

Düngupeyem is a finite state morphological analyser and generator for Mapudüngun, and is based on the algorithms developed by Mans Hulden at his project FOMA (https://code.google.com/archive/p/foma/), an open source application that compiles finite state transducers.

While processing, Düngupeyem applies spelling unification first, analysing next.

The rules for Düngupeyem's performance, and the tags it uses are based on the description and study of Mapudüngun made by Dr. Ineke Smeets and published at her book "A Grammar of Mapuche".

[Ineke Smeets (eds.: Bossong, G., Comrie, B., Dryer, M.) A Grammar of Mapuche. Part VIII, Texts. Berlin, New York: Mouton de Gruyter, 2008]


Ngülamwirinten (Spell checker for Mapudüngun [prototype])

This application is based on Düngupeyem.

We are working to convert it into a useful application, up to this moment it is just a prototype, therefore, it could fail at some point.


IECMap (Corpus exploitation interface for Mapudüngun)

This CQPWeb interface was created by Andrew Hardie (Lancaster University). Our version is adapted and in translation process, even though we offer the original English version also.

The Mapudüngun corpus are tagged with Düngupeyem, while the Spanish and English ones are tagged with FreeLing [License conditions].


ISECMap (Simple corpus exploitation interface for Mapudüngun)

This interface has been completely developed by chandia.net, even though it's also based on the Corpus Query Language (CQP) developed by CWB. It is an attempt to give a simpler access to corpus exploitation for not well trained users who which to exploit data contained in the Mapudüngun corpus.

IECMap and ISECMap share the same corpora.


KMT - Kümewirin Mapudüngun Trapümwe (Spelling unifier for Mapudüngun)

KMT is an orthografical unification tool for Mapudüngun, it's aim is to provide of electronic texts written in Mapudüngun with an uniformed orthography or with the minimal possible variation.

The current version is almost entirely developed in pyhton, a powerful object oriented programming language. Other used codes are html, to enable the user interaction interface by web, and java, implemented at the text editing utility, the bar above the text box. Java allows the command execution by buttons at the web interface.